Index: [Article Count Order] [Thread]

Date:  Sun, 17 Feb 2002 06:13:57 +0900 (JST)
From:  GOTOU Yuuzou <gotoyuzo@notwork.org>
Subject:  [webricken:23] Re: Keep-alive
To:  webricken@notwork.org
Message-Id:  <20020217.061357.68565666.gotoyuzo@kotetsu.does.notwork.org>
In-Reply-To:  <m2eljlttnq.fsf@zip.local.thomases.com>
References:  <m2wuxet4di.fsf@zip.local.thomases.com>	<20020216.171233.26976823.gotoken@notwork.org>	<m2eljlttnq.fsf@zip.local.thomases.com>
X-Mail-Count: 00023

In message <m2eljlttnq.fsf@zip.local.thomases.com>,
 `Dave Thomas <Dave@PragmaticProgrammer.com>' wrote:
> GOTO Kentaro <gotoken@notwork.org> writes:
> > Yuuzou, what do you think about config[:KeepAlive]?  It may halp
> > debug.
> 
> Can you think of a reason why disabling keep_alive would fix it? Is
> there maybe a problem with native Ruby's tcp stack not flushing
> output?

Sorry, I have no idea.
Under Win2k and IE6, I couldn't reproduce the problem.

Instead, I fixed HTTPServer so that the Keep-Alive
connection could be disabled, when :HTTPVersion was
set to 1.0.

Please try the CVS snapshot.

 http://www.notwork.org/ipr/cgi-bin/cvsweb.cgi/src/webrick/webrick.tar.gz?tarball=1

regards,

--
gotoyuzo